Long-term Risk of Breast Cancer Mortality after Local Failure in Women Treated with Breast Conserving Therapy
C. T. Murphy,L. Wang,T. Li,E. I. Obeid,R. J. Bleicher,T. Shaikh,T. M. Churilla,S. B. Hayes,P. R. Anderson
DOI: https://doi.org/10.1016/j.ijrobp.2015.07.327
2015-01-01
Abstract:To estimate long-term risk of ipsilateral breast tumor recurrence (IBTR) after breast conserving therapy and identify subsequent predictors of breast cancer mortality (BCM) after local failure. The study population consisted of women treated with lumpectomy, adjuvant whole-breast radiation plus a boost and known hormone receptor status identified in a prospective database from 1981–2013. Treatment with neoadjuvant chemotherapy was excluded. Characteristics analyzed to predict IBTR risk included: age, T and N stage, grade, lymphovascular invasion, receptor status, margin status, adjuvant systemic therapy (ADJRx) use (defined as endocrine therapy, trastuzumab, or chemotherapy), RT boost dose, and treatment year. Time to IBTR as a component of first failure was defined as in-breast recurrence only, excluding regional nodal failure. For patients with IBTR without distant metastasis (DM), BCM was estimated from the time of IBTR. Chi-square and Wilcoxon tests identified associations between characteristics; univariate log-rank testing on Kaplan-Meier estimates and Cox proportional hazard models (MVA) assessed IBTR and BCM as appropriate. A total of 3 045 women were included. Median age was 58 (range, 20–91), and follow-up was 90 months (range, 2–359) for all patients, and 160 months (range, 14–315) for women with IBTR. For all women, 98.6% had pT1–2 tumors, 74% were pN0, 19% were pN1; 96% had negative margins; 195 were triple negative (TNBC), and 297 were Her2+; 73% received ADJRx. The 20-yr estimated IBTR risk was 16% for the entire cohort (140 pts available for risk assessment at 20 years). In women <50 yo, ADJRx halved the 20-yr risk of IBTR versus no ADJRx (19% vs 38%, P < .001), with a more modest effect in patients >50 yo (8% vs 13%, P < .001). When receiving ADJRx, TNBC had increased 15-yr risk of IBTR when compared to non-TNBC (15% versus 8%, P = .003). On MVA, younger age (continuous), Her2+, TNBC, and no ADJRx were associated with increased risk of IBTR. After excluding pts with concurrent DM, 158 pts had IBTR at first failure, 90% of whom received subsequent mastectomy. The 10-yr BCM was worse in women with IBTR after upfront ADJRx versus none (37% vs 26%, P = .01). A second MVA was performed to predict BCM after IBTR. Younger age (HR = 1.04 95% CI = 1.01–1.07), Her2+ status (HR = 5.03, 1.01–25.2), and positive lymph nodes (LN) at initial diagnosis (HR = 7.9, 2.96–21.2) were associated with increased risk of breast cancer death after IBTR. The overall local failure risk of 0.8% per year in this cohort is mitigated by age, receptor status, and ADJRx. The local control benefit of ADJRx persists up to 20 years after RT, and is greatest for women <50 yo. Despite salvage mastectomy after IBTR, younger women have an increased risk of breast cancer death associated with local failure after initial ADJRx, indicating selection for more aggressive biology.